The Marketing Operating Layer

The average enterprise marketing organization now runs 15 to 50 distinct tools. There's a CDP for customer data, a CRM for relationship management, a DSP for programmatic buying, an ESP for email, an analytics platform for measurement, a content management system, a loyalty platform, a personalization engine, and that's all before you count the AI tools (and recently spun up agents) that have been added in the last two years.

The irony is that more tools have produced more fragmentation, not less. Data lives in silos. Workflows require manual handoffs between systems. Insights generated in the warehouse take 48 hours to reach the activation platform. Campaigns that should be coordinated run in isolation. The tools are sophisticated. The stack, as a system, is dysfunctional.

Now, with the press for employees to learn and adopt generative AI, which has now recently shifted to encouragement that they develop their own AI agents, this problem is replicating itself but the proliferation of agents accessing and utilizing these disparate data sources, it’s compounding the issue and creating even more complexity. We are hearing from CMOs that their team has between 20- and 50 different marketing agents all running simultaneously, and no one knows how they were built and for what purpose. 

This isn't a data problem. Your organization has more data than it can act on. It's a coordination problem and coordination problems require a different class of solution than another point solution.

What is a marketing operating layer?

A marketing operating layer sits above your existing stack, not inside it. It doesn't replace your CDP, your CRM, or your activation tools. It orchestrates them, connecting data, decisions, and actions across the full system in real time.

Think of it the way you'd think about an operating system on a computer. The applications are valuable on their own. But without an OS coordinating memory, I/O, and inter-process communication, those applications can't work together effectively. The OS isn't one more application, it's the layer that makes the other applications composable.

For enterprise marketing, the operating layer performs three core functions:

1. Data unification: Synthesizing signals from across your stack — behavioral data, transactional data, campaign performance, first-party signals — into a coherent, real-time view that agents can act on.
2. Decision orchestration: Evaluating signals against defined strategies and playbooks to determine the right action, at the right time, across the right channel.
3. Cross-system execution: Triggering actions across the downstream tools your organization already has (Braze, Salesforce, Google Ads, your DSP) without requiring manual workflow for each step.

Why "build or buy" is the wrong frame for the path forward

Most enterprise technology decisions get framed as "build versus buy." Build gives you control and customization; buy gives you speed and maintenance leverage. For a marketing operating layer, we believe that neither option alone works for the new agentic AI era.

Pure build means your engineering team is writing infrastructure code instead of building product. It means your AI capabilities are bespoke, expensive to maintain, and isolated from the broader ecosystem of model improvements. It means you're solving coordination problems for one organization without benefiting from learnings across many.

Pure buy, or adopting a vendor's pre-packaged platform, means forcing your organization's architecture, data environment, and workflows into someone else's model. It means your competitive differentiation is bounded by what the vendor has built for the median customer.

The right answer is co-creation, or what Kana calls “build with:” a system built with your architecture, your data, your strategic priorities, but built on a foundation that handles the infrastructure so your team can focus on the strategy.

This is what it means to have an agentic marketing platform that works with your organization rather than replacing what you've built. Your data stays in your environment. Your governance requirements are respected. Your existing stack is orchestrated, not displaced.

What this looks like in practice

At a Fortune 1000 CPG brand, a marketing operating layer means campaign decisions across 15 countries and 8 sub-brands are coordinated through a single intelligence layer.

At an enterprise SaaS company, it means a churn signal detected in Snowflake triggers a retention workflow in Braze within minutes or hours, not days, without a human needing to be involved in every single handoff.

At a mid-market retailer, it means a five-person marketing team can run the personalization and testing programs that previously required a team of 25, because agents handle the orchestration work that used to consume so much of the human team's bandwidth.

The shift that's already happening

The organizations that are winning the transition to agentic marketing aren't the ones that added the most AI tools. They're the ones that built the coordination layer that connects data to decision to action — and then let agents run the playbook continuously.

If your marketing stack is sophisticated but your results still depend on manual orchestration, the investment that moves the needle isn't a new point solution. It's the layer that makes the rest of your stack work together.